Which platforms support Apache Iceberg?

Summary

• Apache Iceberg is supported by major platforms including Databricks, Snowflake, AWS, Google BigQuery, and Microsoft Fabric, along with engines like Spark, Trino, and Flink.
• Key Iceberg features driving adoption include schema evolution, partition evolution, ACID transactions, time travel, and hidden partitioning for petabyte-scale data lakes.
• Databricks treats Iceberg as a first-class format in Unity Catalog alongside Delta Lake and Parquet, providing unified permissions, lineage, and interoperability via the Iceberg REST Catalog API.

Which platforms support Apache Iceberg?

Apache Iceberg has become one of the most widely adopted open table formats for large-scale analytics. Support now spans major cloud providers, query engines, and data platforms. Choosing the right platform means looking beyond basic format compatibility to consider governance, interoperability, and performance at scale.

How Apache Iceberg became an industry standard

Apache Iceberg was originally conceived at Netflix in 2017 to address shortcomings in Apache Hive. It defines how data files, metadata, and snapshots are organized in object storage, enabling ACID transactions, schema evolution, and time travel on data lakes.
Adoption accelerated significantly in 2024 as major providers deepened their Iceberg investments:

• AWS announced specialized S3 table bucket types for native Iceberg support.
• Google BigQuery added native Iceberg table support.
• Microsoft Fabric introduced "Iceberg Links" for access to Iceberg tables.
• Databricks acquired Tabular, a startup founded by Iceberg creators Ryan Blue, Daniel Weeks, and Jason Reid.

According to Gartner, by 2028, 80% of enterprise data warehouses will reside in a lakehouse format, an architecture underpinned by open table formats like Apache Iceberg.

Platforms and engines with Apache Iceberg support

Iceberg's vendor-neutral design means many platforms and engines can read and write Iceberg tables.

Platform	Iceberg support approach
Databricks (Unity Catalog)	Manages Delta Lake, Iceberg, and Parquet as first-class formats under one catalog with unified permissions and lineage.
Apache Spark	Native read/write via the Iceberg Spark runtime library.
Trino / Presto	Built-in Iceberg connectors for federated query across catalogs.
Apache Flink	Streaming read/write for real-time Iceberg ingestion.
Snowflake	Snowflake-managed Iceberg tables with read and write support.
Google BigQuery / BigLake	Iceberg support through BigLake for external and managed tables.
Amazon Redshift	Integrates with AWS services and S3 table buckets for Iceberg workloads.
Microsoft Fabric	Iceberg Links for accessing Iceberg tables within the Fabric environment.
Azure Synapse Analytics	Reads Iceberg tables through Synapse Spark pools.

Additional engines like Apache Hive, Impala, and DuckDB also support Iceberg reads.

Key features driving Iceberg adoption

Organizations choose Iceberg for several technical capabilities:

• Schema evolution: Add, rename, or reorder columns without rewriting data.
• Partition evolution: Change partitioning schemes without data rewrites.
• Time travel: Query historical snapshots for auditing or rollback.
• ACID transactions: Ensure atomicity and consistency for concurrent reads and writes.
• Hidden partitioning: Automatically derive partition values so users write simpler queries.

These features reduce operational overhead and improve query performance on petabyte-scale datasets stored in Amazon S3, Azure Data Lake Storage, or Google Cloud Storage.

How Databricks supports Apache Iceberg

Databricks treats open formats, Delta Lake, Iceberg, and Parquet, as first-class citizens, not bolt-ons. Unity Catalog provides one catalog for all data, managing these formats with a single set of permissions, lineage, and business definitions.

• Managed Iceberg tables: Create tables via Databricks or external engines that support the Iceberg REST Catalog API, such as Spark, Flink, Trino, or Kafka.
• Foreign Iceberg tables: Access and govern Iceberg tables managed by foreign catalogs such as AWS Glue or Hive Metastores.
• Performance optimization: Serverless SQL Warehouse, Photon, Predictive IO, and Intelligent Workload Management deliver performance across Iceberg workloads.
• Orchestration: Lakeflow handles unified pipelines covering both batch and streaming workloads.

Governance, semantics, and lineage are built into the Databricks Platform via Unity Catalog, ensuring one trusted source for every tool.

FAQs

What is Apache Iceberg and how does it work as a table format?

Apache Iceberg is an open-source table format that supports schema evolution, time travel, and hidden partitioning. It maintains atomicity by writing new metadata files for each table change, enabling ACID transactions on object storage.

How does Databricks support Apache Iceberg natively?

Unity Catalog manages Iceberg tables alongside Delta Lake and Parquet with unified permissions, lineage, and business definitions. External engines access managed Iceberg tables via the Iceberg REST Catalog API.

What are the key features of Apache Iceberg that make it popular for data lakehouse architectures?

Schema evolution, partition evolution, ACID transactions, time travel, and hidden partitioning make Iceberg well suited for multi-engine lakehouse environments.

How do you read and write Apache Iceberg tables in a cloud data platform?

Most platforms provide native connectors or catalog integrations. Spark, Flink, and Trino read and write Iceberg tables directly using the Iceberg library or REST Catalog API.

What are the benefits of using Apache Iceberg for large-scale data lake management?

Iceberg delivers ACID transactions, efficient metadata handling, and partition evolution, reducing operational overhead on petabyte-scale datasets.

Which cloud storage systems are compatible with Apache Iceberg?

Iceberg works with Amazon S3, Azure Data Lake Storage, and Google Cloud Storage.

How does Apache Iceberg handle schema evolution and partition evolution?

Schema evolution lets you add, rename, or reorder columns without breaking existing queries. Partition evolution changes the partitioning scheme for new data while leaving existing data in its original layout.

What query engines can natively read Apache Iceberg tables?

Apache Spark, Trino, Flink, Presto, Hive, Impala, and DuckDB can all natively read Iceberg tables.

How do you migrate existing data lake tables to Apache Iceberg format?

Most platforms offer in-place migration utilities that convert existing Parquet or Hive tables to Iceberg without rewriting data files. On Databricks, Unity Catalog manages the lifecycle of both Delta Lake and Iceberg tables, simplifying migration.

What role does Apache Iceberg play in enabling open and interoperable data lakehouse architectures?

Iceberg provides a vendor-neutral table format that any compatible engine can read and write, enabling interoperability across tools and teams.

Build your Iceberg lakehouse on a unified foundation

Apache Iceberg support is now widespread, but platform-level governance and performance optimization separate a production-ready lakehouse from basic format compatibility. Databricks treats Iceberg as a first-class citizen alongside Delta Lake and Parquet, all unified under Unity Catalog with built-in lineage, permissions, and semantics. Explore the Databricks Platform to see how unified governance powers your Iceberg lakehouse.